Building Dynamic, 3D Simulations of Coronavirus Spike Glycoproteins

Assistant chemistry professor Mahmoud Moradi.

Mahmoud Moradi’s work might be the unconventional strategy for developing a coronavirus vaccine.

Only weeks after the virus broke in the United States, Moradi, an assistant professor of chemistry in the J. William Fulbright College of Arts and Sciences, was granted access to Frontera, a supercomputer sponsored by National Science Foundation. It allowed him to build dynamic, three-dimensional simulations of coronavirus spike glycoproteins.

In this new episode of Short Talks From the Hill, a research podcast of the University of Arkansas, Moradi explains how the spike proteins of SARS-CoV-2, the cause of COVID-19, bind to human cell receptors. He then discusses what his simulations have revealed so far and why these findings are important.

“Because the activation of this region itself is important, the conformational, the structural change of this region itself is very important,” Moradi says in the podcast, “there might be ways of designing drugs that stop this region from even being activated. That’s a completely different strategy that the simulations — that we are currently running — might be able to provide some framework for.”

Matt McGowan: Welcome to Short Talks From the Hill, a podcast of the University of Arkansas. My name is Matt McGowan. I’m a science writer here at the university.

Today, I’d like to welcome Mahmoud Moradi, assistant professor of chemistry in the J. William Fulbright College of Arts and Sciences. Moradi is a computational chemist. Using theories from chemistry and physics, computational chemistry is the application of computational techniques and methods, particularly computer simulations, to study and ideally understand the structure and dynamics of molecules. Earlier this year, Moradi received a prestigious $650,000 Early Career award from the National Science Foundation. This award has enabled him to advance his work on modeling the function of proteins at the molecular level. This research will deepen our understanding of disease and improve drug design. More recently as the coronavirus pandemic broke, Moradi gained access to Frontera, a National Science Foundation-sponsored supercomputer and the largest supercomputer on any university campus. Frontera will allow Moradi to build dynamic, three-dimensional simulations of coronavirus spike glycoproteins. These simulations will help scientists understand how the coronavirus binds to human cells.

Before we discuss the coronavirus research, can you talk a little bit about supercomputers? Why are they necessary for your work?

Mahmoud Moradi: The proteins are pretty big in terms of the number of atoms that they have. So if you look at the actual size of them, they’re very small, they are microscopic things, but if you look at the number of atoms that they have… And in physics, we call it the number of degrees of freedom. Every atom has three degrees of freedom. It can go up and down, right and left, you know, it can have three degrees of freedom. So if you consider that, a protein has tens of thousands of those sometimes, and if you also include the environment of protein, because in many of our projects, we have shown this explicitly, how important the environment of a protein is. So if you consider the environment of protein, you can easily get to millions of atoms. And that is the reason that we need supercomputers, to be able to study these systems, because we are dealing with millions of degrees of freedom sometimes, and every degree of freedom, for computational chemistry, means one or more differential equations to solve. So if you want to solve millions of differential equations, of course, you’re going to need the biggest supercomputers in the world.

McGowan: By now most people have seen the high-resolution images of coronavirus spike proteins with the sort of bulbous crowns at the end. What do these proteins do, and why are they important?

Moradi: Absolutely. So, these spike proteins are very visible if you look at very popularized images of the virus, the coronavirus. These are the kinds of spiky things that are coming out of the virus, and they are actually very important. And the main reason for their importance is that this is how the virus recognizes the human cell, it binds to the human cell, and it enters the human cell. So, it is involved in multiple different stages and very initial stages of the infection, basically. So it is very important for us to understand the whole process, to understand how it recognizes the human cell, how it binds to it and the rest. So a lot of people are actually working on spike proteins, and a lot of drugs and vaccines that are being developed are exactly targeting this particular protein because of its significance.

McGowan: So can you tell us what you’re doing to address the problem? What can you say about the molecular dynamics of the coronavirus spike protein?

Moradi: Yeah, so what we are trying to do is, we try to understand spike protein at a level which is not doable using regular computers or regular supercomputers, like the one that we have on campus that we used for other projects a lot. The spike protein is just like any other protein, it has a dynamical aspect to it. And, so far no one has actually studied that. Neither for the current coronavirus, the so-called SARS-Cov2, nor for the previous SARS coronavirus, which is very similar to this one, which is called SARS-Cov. We are interested in both of them, because studying these two proteins, these two viruses and the spike proteins of these two viruses together, allows us to understand why this particular virus is so different. And why is it so deadly or so contagious. And that’s why we are trying to actually study both at the same time, it gives us a frame of reference, when we look at the previous one and the novel one. What we are trying to do with the computational resources that they granted is to look at the dynamics of SARS-Cov2 spike protein and SARS-Cov spike protein. And the reason for that is, as far as I know, no one is actually has published anything on this or has provided any information on this. But I know that several other researchers are probably doing very similar things with this particular protein. Our approach, however, is different. This is using… this project is using an approach that we’ve been developing over the last several years, which is unique to our lab. And it allows us to look at the dynamics on very long time scales, which we think is very important to look at the dynamics on right physiological timescales as well. So, so far, we have been able to look at the microsecond level of structural dynamics of this protein, which is already giving us a lot of information on this protein. And surprisingly, we see that the two spike proteins are very, very different in their dynamics, and this is something that no one has observed before or has reported before. So everybody, so far, is looking at those particular amino acids that are involved in binding to the angiotensin converting enzyme 2, the ACE2, the human receptor. Of course, those amino acids are important, and the fact that that there are some differences between the two virus, the two spike proteins is something that we don’t think is not significant, but at the same time, we have observed there is something probably even more important, which is different between the two. And that’s the dynamical aspect of the two proteins. We’ve observed that the spike protein for the SARS-Cov2, for this novel coronavirus, is a lot more active than what we see in the current structures that are coming from cryogenic electron microscopy. We think it’s actually a lot more active than it is in the SARS-Cov spike protein. And this is based on the microsecond level simulations that we’ve done. The next step that we are working on is to look at the millisecond level dynamics of this protein, which is going to be even more important, because it’s going to give us a bigger picture of what really this protein does.

McGowan: And how could this inform other scientists who are working on a vaccine or medications?

Moradi: The current approach to drug design or vaccine development for this virus involves using the structure of this particular protein, which is, you look at the certain regions of this protein, which is involved in binding the ACE2 receptor. Now, based on the structures that are available, these regions are sometimes inactive and sometimes active. The current approach is basically, mostly revolving on finding some drug that can bind to the binding region of this protein and basically block it from binding to the ACE2. That’s basically a common strategy. What we think is… because the activation of this region itself is important, the conformational, the structural changes of this region itself is very important. There might be ways of designing drugs that stop this region from even being activated. That’s a completely different strategy that the simulations that we are currently running might be able to provide some framework for.

McGowan: Well, Mahmoud, I want to thank you for being here today and talking about your work. It’s so interesting and so important. Thank you.

Moradi: Absolutely. Thank you, Matt.

McGowan: Music for Short Talks From the Hill was written and performed by local musician Ben Harris. For more information and additional podcasts, visit researchfrontiers.uark.edu, the home of research news at the University of Arkansas.

A version of this story also appeared in the University of Arkansas News publication.